Comparison of Ice-Phase Microphysical Parameterization Schemes Using Numerical Simulations of Tropical Convection

Abstract A numerical cloud model is used to evaluate the performance of several ice parameterizations. Results from simulations using these schemes are contrasted with each other, with an ice-free control simulation, and with observations to determine to what extent ice physics affect the realism of these results. Two different types of tropical convection are simulated. Tropical squall-type systems are simulated in two dimensions so that a large domain can be used to incorporate a complete anvil. Nonsquall-type convective lines are simulated in three dimensions owing to their smaller horizontal scale. The inclusion of ice processes enhances the agreement of the simulated convection with some features of observed convection, including the proportion of surface rainfall in the anvil region, and the intensity and structure of the radar brightband near the melting level in the anvil. In the context of our experimental design, the use of three ice classes produces better results than two ice classes or ice-free conditions, and for the tropical cumuli, the optimal mix of the bulk ice hydrometeors is cloud ice-snow-graupel. We infer from our modeling results that application of bulk ice microphysics in cloud models might be case specific, which is a significant limitation. This can have serious ramifications for microwave interpretation of cloud microphysical properties. Generalization of ice processes may require a larger number of ice categories than we have evaluated and/or the prediction of hydrometeor concentrations or particle-size spectra.